Re: [GNUe-dev] explicitfields

[Jason Cater]

Dmitry, 

Sorry for the delay in responding. 

If I recall correctly, explicitfields was added as a "hack" to get certain
summaries to work in GNUe Reports.  I don't think they were meant to do
what you are doing, though they might work with some bugfixes. 

I do see what you're trying to do, and I have needed something similar. We
might want to consider adding something like a  <customfield> tag to
<datasource>. So, your example would read:
 
    <datasource name="dtsHotList" connection="icsinfo" 
                table="user_hotcontact, company"
                order_by="last_update desc">
       <customfield name="contact_info">name||' '||phone</customfield>
    </datasource>

I need to think about this some more though. 

-- 
Jason Cater
GNU Enterprise


On Fri, 23 May 2003 09:58:08 +0400 (MSD)
Dmitry Sorokin <address@hidden> wrote:

> Hello all,
> 
> Question to someone with knowledge of the gnue common.
> 
> 
> I am  using explicidfields in datasource definition in this way:
> 
>    <datasource name="dtsHotList" connection="icsinfo"
>    table="user_hotcontact, company"
>          explicitfields="(name || ' ' || phone) as contact_info"
>          order_by="last_update desc">
> 
> It does not work with current code as "contact_info" field would be
> referenced twice:
> 
> DB005: [Driver:566] SELECT phone,name,contact_info,(name || ' ' ||
> phone) as contact_info,contact_id ...
> 
> 
> ------------------------------------------------------------
> Error: Error while communicating with datasource.
> 
>        ERROR:  Attribute "contact_info" not found
> 
> ------------------------------------------------------------
> 
> 
> I got it working adding some code to remove duplication:
> 
> common/datasources/drivers/DBSIG2/Driver.py
> 
> 
> class DBSIG_DataObject_Object:
>   def __init__(self):
> 
>   ...
> 
>   def _buildQuery(self, conditions={}, forDetail=None,
>   additionalSQL=""):
>     dupfields = []
>     for field_def in self._fieldReferences:
>       field_name = string.split(field_def, " as ")
>       if len(field_name)>1:
>         if self._fieldReferences.has_key(field_name[1]):
>               dupfields.append(field_name[1])
>     for dup in dupfields:
>       del self._fieldReferences[dup]
> 
>     ...
> 
> 
> The question is: is it the right place to fix the problem or it may be
> done earlier to prevent duplication in self._fieldReferences from the
> beginning?
> 
> Thanks in advance,
> dimas
> 
> 
> 
> 
> 
> 
> 
> 
> _______________________________________________
> Gnue-dev mailing list
> address@hidden
> http://mail.gnu.org/mailman/listinfo/gnue-dev
> 


--